| === sikun is now known as sikun-away | ||
| cpaelzer | good morning | 06:59 |
|---|---|---|
| jcputter | hi, i have a question regarding package version in LTS releases of ubuntu. | 07:26 |
| jcputter | i need to install a specific version of openjdk-7-jdk that i need to install, if i browse the repo i find the package however doing apt-cache policy the package is not available | 07:27 |
| jcputter | also when trying to install that version apt complains that it cannot find that package | 07:28 |
| jcputter | any help | 07:28 |
| cpaelzer | jcputter: apt-get policy will get you what currently is most recent and should be used | 07:55 |
| cpaelzer | jcputter: how did you try to install the old version | 07:55 |
| cpaelzer | jcputter: as long as it is in the archive you should be able to try with apt-get install package=version | 07:55 |
| cpaelzer | note that you might have to take care of dependency resoltion on your own in that case | 07:56 |
| cpaelzer | you can list mulitple packages with their versions in the install commandline to do so | 07:56 |
| cpaelzer | jcputter: on trusty e.g. you should still have those three in archive http://paste.ubuntu.com/22891459/ | 07:59 |
| ronator | hi. what would be the best way on ubuntu to switch to systemd-networking after upgrading from ubuntu 14 to 16.04.1? I heard the ubuntu camp decided not to force old installations to switch to new interface naming scheme. I would like to switch and I am looking for a reliable howto. | 08:00 |
| jcputter | i've tried installing the package with apt-get install package=version however apt complains it cannot find the package however i do see it in the repo | 08:10 |
| cpaelzer | jcputter: which release are you on? | 08:14 |
| cpaelzer | jcputter: and what version are you trying to install? | 08:14 |
| jcputter | 12.04 | 08:17 |
| cpaelzer | jcputter: and what jdk are you trying to install? | 08:19 |
| jcputter | openjdk-7-jdk-u79-2.5.5 | 08:36 |
| === _degorenko|afk is now known as degorenko | ||
| cpaelzer | jcputter: so you want this https://launchpad.net/ubuntu/+source/openjdk-7/7u79-2.5.5-0ubuntu0.12.04.1 | 08:46 |
| cpaelzer | that is no more published - as you usually only have the last one in release/updates/security | 08:46 |
| cpaelzer | see at the top of https://launchpad.net/ubuntu/precise/+source/openjdk-7 | 08:46 |
| cpaelzer | jcputter: you can rebuild it from the dsc listed at the first link if you need "just this" | 08:47 |
| cpaelzer | I don't know if there is a "somewhere" to get the old formerly publsihed package (there might be but I don't know) | 08:47 |
| cpaelzer | you find a verson closer than what you want in debian oldstable 7u79-2.5.5-1~deb7u1 - but in general installing forced versionsn and very old things is the start to break your system either by dependenceis or security risks IMHO | 08:49 |
| cpaelzer | smoser: philroche: please take a look at https://code.launchpad.net/~paelzer/cloud-init/fix-mcollect-for-checkers/+merge/302516 | 09:13 |
| === iberezovskiy|off is now known as iberezovskiy | ||
| jamespage | cpaelzer, hey - depending on the feedback we get from the ovs team today, I'd quite like to upload a snapshot of ovs to yakkety | 09:59 |
| jamespage | cpaelzer, my thinking being we can add skip list for know test failures based on arch for the unit test rechecks | 09:59 |
| jamespage | coreycb, ddellav: fixed that xenial keystone/newton failure - we needed to bump freezegun | 10:18 |
| ronator | what would be the best way on ubuntu to switch to systemd-networking after upgrading from ubuntu 14 to 16.04.1? | 10:35 |
| ktt9 | Hello. I have a problem with preseeding ubuntu-server installation. I use repacked installation .iso, but I don't modify squashfs image, instead I choose to store my preseed.cfg on separate image. So the problem is that preseed/early_command in kernel cmdline seem to work AFTER preseed/file since 16.04 server, and not BEFORE, as in earlier versions of Ubuntu. And I use early_command to mount separate medium with preseed.cfg. | 10:39 |
| ktt9 | Am I right about order of preseed/-operations? And if I am, what are my options now? | 10:40 |
| bekks | preseed/early_command is run right before the partitioning. | 11:02 |
| ktt9 | Hmm... I tried to pass preseed/early_command without passing preseed/file and it worked, but of course installation proceeded in manual mode. | 11:04 |
| bekks | what does your preseed/file do actually? | 11:04 |
| ktt9 | preseed/file specifies path to preseed.cfg. And it does all the things. Package installation, language, timezone, keyboard layout. Install is fully automated. | 11:06 |
| bekks | So you could use the preseed/early_command in your preseed/file, right? | 11:07 |
| ktt9 | No, I can't. preseed.cfg is on separate medium, and I used early_command passed to kernel to mount that medium. | 11:09 |
| bekks | ah, I see. ok, maybe you could use another method of deploying it then, like not requiring a mount, but serving it using http, e.g.? | 11:09 |
| bekks | Thats what I do for my preseed/kickstart install, I'm serving all files required using http. | 11:10 |
| ktt9 | That may be an option. I now trying to tell the kernel to mount all media somehow, or to bypass init without repacking squashfs image. Repacking is also an option, btw. | 11:11 |
| ktt9 | Thanks. | 11:11 |
| === ashleyd is now known as ashd | ||
| coreycb | jamespage, qemu 1:2.3+dfsg-5ubuntu9.4~cloud1 is ready to promote to liberty-proposed please | 12:41 |
| coreycb | ddellav, are you working on os-vif? | 13:23 |
| ddellav | coreycb yes | 13:23 |
| coreycb | ddellav, great, thanks. jamespage should be able to upload that to debian for you when it's done. | 13:24 |
| smoser | cpaelzer, bzr is dead | 14:21 |
| cpaelzer | smoser: just found that | 14:25 |
| cpaelzer | smoser: alsready dropped the MP after testing the git | 14:25 |
| smoser | cpaelzer, i'm going to push a "delete everything commit" | 14:25 |
| smoser | with a README that says moved to git. | 14:25 |
| cpaelzer | will surely help to avoid accidents | 14:25 |
| smoser | sorry for the noise, cpaelzer | 14:25 |
| cpaelzer | smoser: I didn#t know you have opened up the git already til the mail | 14:25 |
| smoser | i just sent an email to ubuntu-server and cloud-init mailing list that says 0.7.7 is released. | 14:25 |
| cpaelzer | smoser: I'll move over my bzr branch for the apt things then | 14:26 |
| cpaelzer | smoser: yeah that is the mail that got me going with the MP delete | 14:26 |
| ikonia | brz is dead ? | 14:28 |
| ikonia | so has launchpad stopped using it ? | 14:28 |
| smoser | no, ikonia i meant only in the context of cloud-init | 14:28 |
| smoser | bzr is still alive and fully supported on launchpad, but we've moved cloud-init over to git. | 14:29 |
| ikonia | ah | 14:29 |
| === ivyyyy-brb is now known as ivyyy | ||
| Xin | hey all | 14:39 |
| Xin | I want to do some thinclient style stuff - but I want to do it more like citrix than vnc | 14:39 |
| Braven | hello | 14:40 |
| Xin | Essentially what I was thinking was to use many x11 sessions with a predefined resolution for each virtualized app instance | 14:40 |
| Xin | and then I remotely connect to all the x11's from a remote x11 | 14:40 |
| Xin | but I have no idea if that's right | 14:41 |
| Xin | or if there's an easier way | 14:41 |
| Xin | lol | 14:41 |
| Braven | Has anyone been answering? | 14:42 |
| Braven | hello | 14:53 |
| MASM | some one want to help me with my problem of RAID 1 (mirror) http://askubuntu.com/questions/809823/error-raid1-not-unsynchronized-a-degradedarray-event-had-been-detected-on-md-d | 15:27 |
| MASM | Some one want to help me with my problem of RAID 1 (mirror) http://askubuntu.com/questions/809823/error-raid1-not-unsynchronized-a-degradedarray-event-had-been-detected-on-md-d | 15:35 |
| ivoks | MASM: are md1 and md3 mounted? | 15:38 |
| MASM | I don't know, i think md1 is mounted | 15:55 |
| MASM | sorry for last a lot, I was looking for information | 15:56 |
| MASM | ivoks: I think md1 is mounted, http://paste.ubuntu.com/22922653/ | 15:56 |
| ivoks | MASM: so, in your comment you say you installed on sda | 15:57 |
| ivoks | MASM: but mount shows that your / is on md1 | 15:57 |
| ivoks | MASM: and i can assume that md3 is a PV used by vg00 VG | 15:58 |
| MASM | I think i'm in md1 that has /dev/sda | 16:00 |
| MASM | and i installed all programs in it | 16:00 |
| MASM | and raid are not sincronized | 16:00 |
| MASM | and /dev/sda has some read failure | 16:00 |
| MASM | that show in the test | 16:00 |
| MASM | 20 % ~ | 16:00 |
| cpaelzer | jamespage: are coming to host the HO? | 16:00 |
| cpaelzer | jamespage: since I joined first it passed all the "authoenticate others" to me | 16:01 |
| ivoks | MASM: output at http://askubuntu.com/questions/809823/error-raid1-not-unsynchronized-a-degradedarray-event-had-been-detected-on-md-d shows that sda is not part of raids | 16:02 |
| ivoks | md1 : active raid1 sdb1[1] | 16:02 |
| ivoks | md3 : active raid1 sdb3[0] | 16:02 |
| ivoks | no sda there | 16:02 |
| MASM | it means | 16:03 |
| MASM | that i'm in sdb? | 16:03 |
| ivoks | it means you are in md1 | 16:03 |
| ivoks | and md1 currently only has sdb1 in it | 16:03 |
| ivoks | you have two raids; md1 (sdb1 and something that's missing) and md3 (sdb3 and something that's missing) | 16:04 |
| MASM | And what i need to do? | 16:06 |
| MASM | renplace sda ? and config raid software? | 16:07 |
| MASM | mdadm? | 16:07 |
| ivoks | MASM: judging by your smart tests, your sda might die at some point | 16:07 |
| ivoks | MASM: you can still add it back to raids, but expect to be buying a new disk at some point | 16:07 |
| ivoks | MASM: mdadm --manage /dev/md1 --add /dev/sda1 | 16:08 |
| ivoks | MASM: mdadm --manage /dev/md3 --add /dev/sda3 | 16:08 |
| ivoks | that will add those two partitions to those two raids | 16:09 |
| ivoks | MASM: and then you can check /proc/mdstat to see progress of syncronization | 16:09 |
| ivoks | note; whatever you have on sda1 and sda3 will be deleted | 16:10 |
| ivoks | but judging by your /etc/mdadm/mdadm.conf, this is what your raids were like before sda got kicked out | 16:11 |
| ivoks | (and it will get kicked out again) | 16:11 |
| ivoks | so, start buying a new disk :) | 16:11 |
| MASM | ivoks: i will not lost information with this commands?, what is the reference to copie information, what is the reference to copie information? | 16:11 |
| ivoks | lost information? | 16:12 |
| ivoks | you had raid1 (md1), made of sad1 and sdb1 | 16:12 |
| ivoks | those partitions (sda1 and sdb1) were identical | 16:12 |
| ivoks | linux kicked sda1 out of md1 because the disk is dying | 16:12 |
| ivoks | it did the same with md3 and sda3 (for the same reason) | 16:13 |
| ivoks | you can re-add sda1 and sda3 to their raids, but if the disk is bad, it will get kicked out of the raid again | 16:13 |
| MASM | ivoks: ok, thank, i will send a mail to server support to specify that sda is dying | 16:14 |
| MASM | and when i they change the disk, the server automatic sincronized the disks? | 16:15 |
| ivoks | no | 16:15 |
| ivoks | because new disk will have no partitions | 16:15 |
| MASM | ivoks: or i need to copy the partitions, and sincronize it? | 16:15 |
| ivoks | you will have to create partitions | 16:15 |
| ivoks | and then re-add those partitions in md1 and md3 | 16:16 |
| MASM | ivoks: Thanks for your help ivoks, I really appreciate your help | 16:18 |
| ivoks | np | 16:18 |
| MASM | ivoks: just one more question | 16:18 |
| ivoks | MASM: keep in mind that support probably won't be able to do anything with 'replace sda' information. they will prefer serial number of the disk (see your smartctl output) | 16:19 |
| ivoks | S1D9YP5D | 16:20 |
| MASM | ivoks: for resuming, i'm in md3 that has sdb, that the server is using, sda is kickout, all my information is on sdb, right? | 16:20 |
| ivoks | MASM: right | 16:21 |
| MASM | ivoks: and when they replace it i need to copie the partition, and sincronized it again adding the disk, but beffore i need to delete the sda and add the new disk? | 16:22 |
| MASM | and RAID1 will mirror sdb to the new dev? | 16:22 |
| ivoks | MASM: https://www.howtoforge.com/replacing_hard_disks_in_a_raid1_array#-adding-the-new-hard-disk (just notice that they replaced sdb) | 16:23 |
| ivoks | so, don't just copy-paste | 16:23 |
| ivoks | you would do: | 16:23 |
| ivoks | sfdisk -d /dev/sdb | sfdisk /dev/sda | 16:23 |
| ivoks | mdadm --manage /dev/md1 --add /dev/sda1 | 16:23 |
| ivoks | mdadm --manage /dev/md3 --add /dev/sda3 | 16:23 |
| ivoks | mkswap /dev/sda2 | 16:24 |
| ivoks | swapon -a | 16:24 |
| ivoks | *after you replace sda with new disk* | 16:24 |
| sarnold | ivoks: coool, thanks :) I haven't taken the time yet to learn how to feed, care for, and nurture my mdadm raid, hehe | 16:25 |
| MASM | ivoks: Thanks a lot men, I really appreciate your help | 16:25 |
| MASM | ivoks: really !! | 16:25 |
| ivoks | np guys :) | 16:26 |
| === rickyh is now known as rickyhewitt | ||
| === rickyhewitt is now known as rickyh | ||
| === degorenko is now known as _degorenko|afk | ||
| rbasak | soren: would you mind handing over ~ubuntu-server-qa ownership and admin to ~ubuntu-server-dev please? We'd like to reuse this team. | 16:48 |
| solarce | I am hoping to get in contact with someone who's involved in building and release the official GCE ubuntu images. we (travisci) base our build environment GCE images off the official Ubuntu ones, we do a packer + chef build process to customize it. it seems like the setting to disable ipv6 by default in the GCE VMs was lost in an update to the images, at least for Trusty. But I'm not sure where because | 16:51 |
| solarce | we just went from a 2015 release toa 2016 release yesterday, we were based on ubuntu-1404-trusty-v20150909a before and now we're based on ubuntu-1404-trusty-v20160627 | 16:51 |
| solarce | rcj: Odd_Bloke: from what i can gather, you two may be some of the folks i'd want to talk to? | 17:03 |
| xnox | gaughen, i think you want to talk to solarce ^ | 17:09 |
| xnox | "<solarce> I am hoping to get in contact with someone who's involved in building and release the official GCE ubuntu images. we (travisci) base our build environment GCE images off the official Ubuntu ones, we do a packer + chef build process to customize it. it seems like the setting to disable ipv6 by default in the GCE VMs was lost in an update to the images, at least for Trusty. But I'm not sure where because" | 17:09 |
| xnox | "<solarce> we just went from a 2015 release toa 2016 release yesterday, we were based on ubuntu-1404-trusty-v20150909a before and now we're based on ubuntu-1404-trusty-v20160627" | 17:09 |
| gaughen | solarce, that was changed due to support for lxd. Odd_Bloke here can probably do a better job giving details, and advising on a path fwd | 17:13 |
| Odd_Bloke | solarce: Yeah, we were disabling it on GCE per their recommendations. | 17:14 |
| Odd_Bloke | solarce: However, lxd uses link-local IPv6 to provide networking to containers. | 17:14 |
| Odd_Bloke | solarce: So once we had a piece of software that was being broken by that default, we made the decision to back that particular modification out. | 17:15 |
| Odd_Bloke | solarce: It sounds like you've worked out how to replace it for your specific use case though. :) | 17:15 |
| === alexisb is now known as alexisb-afk | ||
| === InfoTest1 is now known as InfoTest | ||
| dr4c4n | hey everyone, some help here, say you have a full ubuntu server setup installed as a hypervisor running kvm, which are the important files besides fstab and /etc/networking that you would move to an alternate installation? | 17:21 |
| solarce | Odd_Bloke: we do have a fix we're rolling out | 17:22 |
| solarce | Odd_Bloke: are the tools you use to build the GCE images available for me to look at? | 17:24 |
| sarnold | dr4c4n: libvirt configurations, if you're using that to manage qemu/kvm.. | 17:25 |
| dr4c4n | sarnold: when I did the installation, I just ran default install, and added virtualization on the menu | 17:26 |
| dr4c4n | it installed correctly, I think I added virt-manager to the management box which is external to the hypervisor server | 17:26 |
| === iberezovskiy is now known as iberezovskiy|afk | ||
| arooni | hey folks; i'm running ubuntu 14.04 LTS server on a VPS. i'd like to upgrade it to 16.04; any considerations i must make before doing so? i'm running a few wordpress/woocommerce sites off it. | 17:41 |
| sarnold | arooni: there's no php5 on 16.04 LTS, be sure your applications can cope | 17:41 |
| arooni | ah ha. | 17:41 |
| arooni | so no way to put php5 on it? | 17:42 |
| solarce | arooni: if you can swing some downtime, stopping the VPS and making a snapshot before you upgrade is a good path | 17:42 |
| OerHeks | arooni, vps-vendors tweak their images heavily, ask them for a fresh 16.04 | 17:42 |
| sarnold | I understand ondrej's got a ppa with it, but that's probably it | 17:42 |
| arooni | hmm.... cool | 17:42 |
| arooni | so you would NOT reccommend the upgrade? | 17:43 |
| OerHeks | arooni, indeed. ask your vendor too | 17:43 |
| sarnold | there's no rush, 14.04 LTS will be supported for a few more years, and maybe you deploy the new one when apps are prepared to handle it | 17:43 |
| arooni | is there a compelling reason to be on 16.04 over 14.04? | 17:55 |
| arooni | besides the longer window of support ? | 17:55 |
| OerHeks | you can answer that yourself with these releasenotes https://wiki.ubuntu.com/XenialXerus/ReleaseNotes | 17:56 |
| OerHeks | LXD 2.0/ libvirt 1.3.1 or perhaps php7.0/mysql5.7 | 17:57 |
| solarce | Odd_Bloke: thanks for the info | 18:12 |
| beisner | coreycb, are you able to access https://bugs.launchpad.net/bugs/1611123 ? | 18:27 |
| ubottu | Error: launchpad bug 1611123 not found | 18:27 |
| beisner | ha! that's why i ask. it's the bug in the commit msg for qemu liberty, and our tooling refuses to promote it to proposed since it can't do bug foo. | 18:28 |
| seeeb | Hi guys, I have a couple of scripts in /etc/init/ to start/stop something and I need that to start on boot. Currently is not doing it and there is nothing about them in /etc/init.d what technique should I use to make them autostart? Thanks | 18:54 |
| sarnold | seeeb: what release are you using? | 18:56 |
| seeeb | 14 | 18:56 |
| sarnold | seeeb: /etc/init/ is for upstart configurations; /etc/init.d/ is for older sysv-init style initscripts. If you've got _scripts_, then it's probably best to put them into the /etc/init.d/ directory and use update-rc.d to create all the symlinks that are used for the sysv-compatibility mode | 18:57 |
| sarnold | seeeb: if it's an upstart configuration instead, then you get to debug why it isn't working :) hehe | 18:57 |
| seeeb | hehe | 18:57 |
| seeeb | I think is upstart yeah, so they should already be starting up you think? | 18:57 |
| sarnold | probably; can you pastebin the file? someone might be able to spot a problem | 18:58 |
| seeeb | sarnold: they start (and even restart alone if killed) but they don't do that on boot | 18:58 |
| seeeb | ok let me do that.. | 18:58 |
| sarnold | what's the filename? I think the /etc/init/ directory requires files to be named with .conf at the end.. | 18:59 |
| seeeb | they are | 18:59 |
| seeeb | Here is /etc/init/staging-app-admin.conf for example https://gist.github.com/sebastianconcept/abbe9d7720224ff17e47434a0888038b | 19:03 |
| seeeb | sarnold: | | 19:03 |
| sarnold | seeeb: hmm all seems fine; does this work as expected? | 19:05 |
| sarnold | start on started mongodb and runlevel [2345] | 19:05 |
| sarnold | seeeb: .. e.g. does mongodb start? :) | 19:05 |
| seeeb | mongo starts alone by itself on boot | 19:08 |
| seeeb | but the ones that are like this one does not | 19:09 |
| sarnold | seeeb: hrm, I never made 'complicated' upstart configurations.. that surely looks normal enough though | 19:09 |
| sarnold | seeeb: flailing around, I think maybe try removing the "and runlevel [2345]" bit | 19:10 |
| seeeb | right, trying that | 19:10 |
| sarnold | seeeb: ah, I forgot my #1 debugging step -- is there anything in the logs? /var/log/upstart/* or /var/log/syslog or ... ? | 19:12 |
| seeeb | cool, will check but now need to wait the reboot heh | 19:13 |
| jge_ | Hey all, good afternoon. I'm trying to extend a VG using LVM but when I try creating a logical partition it says "No free sectors available" | 19:15 |
| jge_ | any ida? | 19:15 |
| jge_ | idea* | 19:15 |
| jge_ | :\ | 19:15 |
| seeeb | lots of logs sarnold! there is an exception about failing to connect to mongo, maybe it needs to start a bit delayed. Do you know if that's doable with upstart? | 19:21 |
| sarnold | seeeb: I -think- the solution to that is along these lines: http://upstart.ubuntu.com/cookbook/#signals | 19:23 |
| sarnold | seeeb: .. you'd have the mongo configuration emit something like mongo-started and have this one 'start on mongo-started' | 19:23 |
| seeeb | giving it a try to that idea.. | 19:24 |
| === alexisb-afk is now known as alexisb | ||
| arooni | is it a huge security hazard to be running passwordless sudo on my VPS server? | 20:59 |
| sarnold | arooni: it does mean that your user account is equivalent to root | 22:06 |
| sarnold | arooni: if you treat it that way, that's fine; if you don't show it the proper respect it could be Big Trouble | 22:06 |
| teward | arooni: instinctively: yes, it's a huge security risk, for the reasons sarnold said | 22:35 |
| teward | but if your VPS server can only be reached through a VNC / Serial Console perpetually, and you don't need SSH on it, then maybe. That said, you run a lot of risks still either way | 22:35 |
| === sikun-away is now known as sikun | ||
| solarce | arooni: only use key authentication for ssh is important, adding two-factor authentication to ssh is a good idea | 23:33 |
| teward | ^ that | 23:36 |
| solarce | i am a big fan of https://duo.com/docs/duounix | 23:37 |
| solarce | use it protect anything that exposes ssh to the internet | 23:37 |
| teward | ^ also that, I use Duo on all my servers' SSH, key auth or not | 23:38 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!